Building a Scraping Technology Business in War-Torn Ukraine – Oleg Kulyk of ScrapingAnt.com

Sushant Misra of TrepTalks: Hey there entrepreneurs. My name is Sushant and welcome to Treptalks. Today, I’m really excited to welcome Oleg Kulyk to the show. Oleg is the founder of Scraping Ant. Scraping Ant offers a web scraping API that allows users to extract data from websites without getting blocked. The platform handles challenges such as rotating proxy servers, headless browsers, CAPTCHA and CloudFare.

And today I’m going to ask Oleg a few questions about his entrepreneur journey and some of the strategies and tactics that he has used to start and grow his business. So, Oleg, thank you so much for joining me today at Treptalks. Really, really appreciate your time. Hello, Sushant.

Oleg Kulyk of ScrapingAnt.com: Hello, dear entrepreneurs. It is nice to be a guest at your show.

Sushant Misra of TrepTalks: Thank you so much. So, you know, we were just before we started this, we were just talking that you are joining [00:01:00] right now from Ukraine. And, you know, right now, uh, most people kind of are aware there’s a war going on in Ukraine, Russia, Ukraine situation. And your kind of located in the heart of it, you know, in Kiev.

Um, so maybe, uh, you know, just give us a little bit of a. background on what is it like, uh, working in Kyiv right now, running a business out of Kyiv. Um, and how are you feeling right now?

Oleg Kulyk of ScrapingAnt.com: Yeah, I would be happy to. So, to be honest, the first year of Russian invasion in Ukraine was really tough. So, we lived under the constant shelling at least the first months of it.

Russian forces were near… My place and due to the family reasons, I was not able to leave my city. So I constantly heard of they are shelling, and I’ve heard that shelling [00:02:00] back from the Ukrainian forces and it was the like 24 hours per day and like it was really hard in terms of business for sure because We wasn’t, we weren’t able to operate properly.

So, the best we could do is to provide the general support of all our customers to communicate with them and to explain the situation. So, we have. And no way to do new features to like to expand our business, to provide more sales. We just had to keep our existing clients and continue them to operate their businesses because they are relying on our data and our service.

But I guess we did it pretty fine as we had no churn. At the end of this, of that year, but it was really hard in terms we had [00:03:00] no electricity, we had no water, sometimes no internet at the end of the year, and like in winter, but we overcome it. And. Now we have the supply of water, of food, of, uh, some, uh, energy banks, like, batteries, some backup internet supplies.

So now we can operate, I guess, even better than before, before the invasion. But still, it’s hard because this is still going on. The invaders… The Russians are still in our country, we still can hear the air radar sirens, it’s still not the best place to grow your business. It’s not that safe. It’s harder to collaborate with other businesses because not all the businesses stayed in [00:04:00] Ukraine.

So yeah, it’s harder in case of, in scope of the… in person meetings, but we are trying to overcome it by doing it all online. And yeah, we are mostly operating from home and starting from this COVID 19 times, we, I guess all of us get used to it. So yeah, it’s something that similar, there is some similarities between the situations, but still, yeah, for sure.

It’s hard.

Sushant Misra of TrepTalks: Yeah, I mean, it’s uh, I mean, I’m glad to hear things are a little bit better now But I think when you are in a war situation thing can probably change pretty quickly. So, there’s always that kind of risk You know, no matter how normal things look, you know, people can bomb, you know in the city anytime like there’s no certainty there, right?

But I’m very interested to learn about your [00:05:00] business Scraping Ant, um, and your entrepreneurial story. So maybe you can share a little bit about your background and what kind of, how did you get the idea for Scraping Ant and what motivated you to start this business?

Oleg Kulyk of ScrapingAnt.com: Yeah, sure. So, the whole, my entrepreneurial like story probably started in my childhood.

So, my parents. motivated me to start doing something by my own because starting from 90s Ukraine wasn’t in a good place because of the USSR and USSR collapsed and all those countries started to grow. And to rise by their own. So, there was, were a lot of problems my parents need to have to overcome. And they taught me to do this, uh, [00:06:00] entrepreneurial and all like entrepreneurial ventures and businesses as they started, uh, a small grocery shop in the village near me.

hometown. So, they provided good goods for people in those village, in that village. And they taught me some basic, some basics of, um, how to do business. What’s the margins, how to calculate profits. And they allowed me to sell some school utensils for before the school year starts. So that money was my first.

money that I earned by my own. So, starting from that, I tried several ventures. I guess most of entrepreneurs tried some stuff like [00:07:00] drop shipping or trying to sell goods, but then COVID 19 started. And we all were closed at our homes and friend of mine also lost a job and we’re looking for a new one. And his experience in data extraction and data gathering was like really huge.

And we came up with this idea to try it out, to create some tool that would help people to solve this problem, to overcome the… Detections block and to test this out, we published this tool at API marketplace. It goes rapid API, and we just started it and published it to check the traction. How good is it, whether people interested in it or no.

And we got pretty good [00:08:00] response from first customers, from first people who tried it. We didn’t spend any cents to promote it because it was a marketplace and it handled all the payments and so on, so on. So, we decided to After that, we decided to, uh, started like full growing business out of it. So, we registered the company, uh, and we started our self-hosted API, self-hosted, uh, website that provides all those services, because it’s really hard to market something.

That you don’t have control of when it’s placed at somewhere at some marketplace. You don’t know the conversion rate. You cannot like place some like links. You cannot run some advertisement companies and so on. So we decided to do it at our [00:09:00] own platform. And yeah, it was the start of this project, Scraping Ant.

Sushant Misra of TrepTalks: So, your, your co-founder is the one who’s the technical one. You are more of the businessperson. Is that correct? Or are you, you also are involved in like the technical coding aspects of this business? Yeah, we

Oleg Kulyk of ScrapingAnt.com: are two technical founders. So, we, both of us like have 10 plus years of, uh, Software development experience, but he had more experience in data extraction, and I was working in different areas, but this one was something new to me and I decided to learn more from him.

So, both of us. Our technical, but, uh, the, [00:10:00] my, oh, but we decided to split responsibilities that way that he would be responsible for all the technical side. And since I had some experience in sales and communication with people, uh, we decided that I would be the salesperson and marketer and, uh, creator of the website and all like one-man bands.

For the marketing and he would take care of the technical

Sushant Misra of TrepTalks: side. I think, I think that makes a lot of sense, uh, for sure. Um, can you talk a little bit about your product? So, what is, what is the product? What does it do? So, I mean, of course, from the, uh, name, it’s a scraping API, but can you translate in like layman terms?

What does it allow people to do? Um, and [00:11:00] who are kind of the main? people that can use this API. Um, so maybe talk a little bit about your product and the main value propositions. What is, what is the benefit of using this product? Yeah, sure.

Oleg Kulyk of ScrapingAnt.com: So, when you need some data from the internet, you’re probably would go to Google and type some search combination, search word, and you will get.

search result. Results open probably one of the first links and observe the information. Still, uh, this process should be automated when you’re doing this at scale. So, because it’s really hard to do such amount of data processing by human. Still, websites…

are not like websites don’t like when you access them [00:12:00] by some automation because they think that it could be a treat for them. You might want to create a denial of the service attack, or you want to do some malicious actions, so they protect themselves. From this automated, uh, access as well as some of the websites could not be accessible from your region but could be accessible from another one region.

And before this, uh, web scraping API concept appeared at the internet, uh, data extraction companies had to combine, uh, proxy IP addresses, create some. Uh, anti-detection systems, create something that would allow them to overcome, uh, cloud flare capture. We [00:13:00] decided not to create some really smart and technical advanced systems, but we got it in a different way.

We decided to act the same as the real human would do. So, in our system, we are running. Real browser and this real browser observing web pages as it would do a real person. So, we don’t trigger the suspicion and those detection system as well as we allow us to do it in parallel for thousands of websites.

So, when you need to get data from let’s imagine 1000 websites, you can do it with. Probably two lines of code and you would get it like in 30 seconds. So, we don’t, [00:14:00] we allow people to get all those data pretty easily without worrying about infrastructures, infrastructure, uh, servers, but by just calling the API endpoint and getting the needed data.

And we recently released it. Pretty cool feature that uses AI to extract data out of HTML. So, you might know that websites are based on HTML, but when you are extracting data from web, you probably don’t need it. You need the phone numbers, you need price, you need product name, and extracting this data out of HTML could be a real pain, so we created.

AI that extracts it for you. So, when you’re calling our API, you just get in the data you need and you’re specifying this data by just [00:15:00] writing what you need. And you can write by product name, product, the price, and you would get it automatically extracted for A really, really affordable.

Sushant Misra of TrepTalks: cost. That sounds very, very interesting.

I mean, everybody, today we are living in a data centric world, and everybody wants to use data. There are so many different applications, but now for AI, of course, if you want to train a model, you need a lot of data to train the model as well. Um, so you mentioned that, you know, a person can write a couple of lines of code, and they’re going to specify thousands of websites and your system will go and scrape data from thousands of websites.

Um, Does the provider have to, does the user have to provide the links of thousands of websites or all the websites that they want to scrape? Or is it that, let’s say that a e commerce use case, [00:16:00] right? You know, someone wants to do some product research or category research from Amazon. How would they do that using your API?

Um, do they have to first provide like a list of all the web pages or websites that they want to scrape the data from? So, can you kind of walk through at least like one use case of how somebody is using your product? Yeah,

Oleg Kulyk of ScrapingAnt.com: sure. So, the most interesting use case, uh, as I think is when you need to extract.

Contact data for the third, their marketing campaign. And I think Google is the best tool for this. So, you can extract data. You can extract website URLs. You want to get information from by scraping Google search output for a specific keyword. For example, you [00:17:00] can look for, uh, uh, companies that works with, uh, let’s say, CEO, CEO.

Marketings and, uh, you are looking for them and you, as a result, you are getting 10 pages of Google search results. And each of those have their unique layout, their unique specifications. Some of them uses dynamic layouts for, and modern technologies to display data. And as we are using browser and AI, you don’t need to worry about all of those.

You just need to go scrape Google and URLs. for the specific, uh, keyword. And then you re you have all those, like imagine thousands of web pages you can get exact info from, and you can just type that [00:18:00] you need phone number, contact name, contact email, and you would get it.

Sushant Misra of TrepTalks: Wow. That’s, that’s awesome. So, if you think about some of these, um, customer data websites that are out there that charge thousands of dollars.

To access their contact database and which is kind of like a broad database with your API, a person can go and just search a term like, you know, marketing agency in Toronto or marketing agencies in New York. And all of the thousand results that they get of the different marketing agencies, your system is going to be able to go on those sites, extract the company, uh, information or, you know, uh, information of the contact, you know, phone number, email address, things like that, and then create a table and provide it to the user.

And what kind of. [00:19:00] Does a person need to be a technical person in order to write code to, to be able to do this, or is this more very straightforward? So,

Oleg Kulyk of ScrapingAnt.com: unfortunately, yes. So, let’s say it this way. So, unfortunately, yes, in case you won’t reach us via any of our contact points. So, we are really Happy to conduct a good relationship with our clients and we want to help them to solve their prob their problems.

So, the integration with our product itself is not that hard and. To really for the use case you mentioned before, and we’ve discussed it before, like extracting contact data, you would probably need not two lines of code, but around six lines of code, which is not that bad. And, uh, we are happy to provide. [00:20:00]

These six lines of code, we don’t charge for that. And we don’t have like specific restriction that if you don’t pay us, we don’t give you the help. We’re always happy to help our clients as bringing the value is one of our main values in the company. We want to, we want to help people and we want to bring the value.

And if we would do that, probably we would have like, we would have the revenue and so on, so on, but conducting a good relationship and helping people is a key for our success. So it’s better when you have someone that could help you with the technical side of the integration. But if no, we would be happy to help people to do.

Sushant Misra of TrepTalks: that.

So then what is your pricing strategy? What does your pricing look like? How do you charge? Um, the number of websites that need to be scraped, how, how exactly is your, does your pricing work? [00:21:00] You said that it’s reasonably, it’s reasonably priced, right?

Oleg Kulyk of ScrapingAnt.com: Yeah. So, uh, we had a fight with the second founder of Scraping Ant regarding the presubscription at our service.

So, most of our competitors are providing free trial. That are just to try their service and then you have to pay, and it leads to. Some malicious actions and multiple account registering and other problems. And a friend of mine, Andrew, who is a, like a founder of scraping and forced me to keep our current free subscription plan, which includes, um, 10, 000 of API credits, it’s let’s call it it’s ours.

Or internal currency. So, in the simplest case, this amount [00:22:00] allows you to get HTML data out of 10, 000 of websites monthly. So, it’s fully featured plan, it doesn’t limit, we only limit the concurrency, so you cannot do it in a thousand parallel requests, you can do it one by one. But… We are, we don’t limit any other features.

You can use all the countries you can use browser. So, in the simplest case. You would be able to extract one, 10, 000 of web pages. Still, if you would need to use some more advanced, uh, scraping capabilities, it would require you to pay a bit more for. One particular request and from now, for now, our AI extractor is still in a better version.

So, we don’t have a fixed price for, for the request, [00:23:00] but it depends on the size of the website web page. If it’s really big, it would cost more rather than in case some website is smaller. So it would be cheaper. But. It’s free. So, we have a fully featured free plan and the motivation behind this, that forced for, from the other founder, that there could be some enthusiasts and small companies that cannot afford, uh, like paying for.

bigger plans so they could start with us, and we could help them to grow so we can grow together and giving something would help us to then get something back, which could be a good strategy. So yeah, I cannot insist on something opposite. So we decided [00:24:00] to keep our free plan and operate this way. So yeah, it starts.

And from a free plan, we have three tiers, which includes, uh, like different price per, uh, one API credit. And we’re also happy to provide any sort of customization for plans. So it really depends on the particular business who want to integrate with us. But we are aiming to work with businesses. So yeah, it’s all about the business needs and we are pretty flexible in scope of what we can provide and what the pricing could be for a particular use case.

Sushant Misra of TrepTalks: So basically, your product, there’s like, to me, it sounds like there’s two components to it. One component is just the scraping and that gives you HTML code. And then the second component is taking that HTML code, utilizing AI and kind of giving it more human readable structure. Is that, would that be a correct statement?

Oleg Kulyk of ScrapingAnt.com: Yep. [00:25:00] Yep. That’s totally correct. So when you are making AI enabled requests, The beginning of this request, um, separated to, to two parts. The first one is extraction of data itself, HTML data. And the second part is, uh, converting this data to human readable format. And so we are planning to get away from this unpredictable price.

It’s still would not be that high, but still, it’s harder to predict how much you would need those CPI credits for your use case. So we are going to finish the beta, the beta testing for this feature and stabilize the price at some constant point. So it would be easier for our customers to estimate their needs.

So

Sushant Misra of TrepTalks: who are your customers? Like what kind of customers do you usually get? Um, are these usually more like [00:26:00] new startups, you know, founders of new startups who are kind of, you know, starting a new project and want the data for that? Are these mostly, you know, companies who are utilizing AI? So they’re looking for, you know, training data from the web.

Um, are these like bigger companies who, who are kind of. Um, utilizing data for competitive purposes or, you know, other things. So what kind of customers do you usually, uh, come across? Do you have like any certain industry or category that, you know, you see more often?

Oleg Kulyk of ScrapingAnt.com: than others? So in terms of, uh, business niches that are using our service, it’s really hard to tell because we, we are trying to communicate and we are.

Always happy to communicate and answer the questions, have a small talk with our clients and so on, but [00:27:00] we don’t insist on this and we don’t force people to receive spam from our emails and so on. So we are friendly at that site. So, it’s really hard to tell, because not all of them are using our service with the business domain, they are connected to.

So, still, I think that the main use case. Is e commerce scraping so we have a lot of requests to get information from amazon from eBay and from platforms that allow selling goods like etsy. So most of our clients Not let’s say not most of our clients, but a lot of requests are going to those websites still As you mentioned before, data and using data from internet is something that could be useful for any business. [00:28:00]

So, someone is interested in, interested in getting data from Google and there are tools that are integrated with our API and allows to get data from Google, human readable data from Google pretty easily. Some of the businesses would like to get a stock market price. Some of businesses monitoring cryptocurrencies using our service.

So, and I think that’s, it’s the. Main problem with marketing main, not the problem, but the main, um, uh,

Sushant Misra of TrepTalks: challenge.

Oleg Kulyk of ScrapingAnt.com: Yeah. Main challenge. You’re right. So, uh, too, because it’s hard to find the niche, one particular niche as you. Could be helpful for all of them. So we are trying to, uh, connect with people and [00:29:00] find as much as possible niches that we could be useful for.

And we don’t focus on one particular niche and when we are trying to. Spread and communicate with different businesses. We are receiving pretty good responses, pretty good feedback. And some of the companies haven’t even think that data could bring them that sort of advantage and cost saving.

Sushant Misra of TrepTalks: Who are your competitors?

I’m assuming that there’s probably other data scraping services and sites, uh, tools that are out there that can perform the same task. Uh, but to me, it seems like. Where you are different is kind of the approach that you take in terms of, um, how you script the data. Um, do you have any direct competitors in the market?

Um, and, um, what is the difference? Like if somebody is… [00:30:00] You know, an entrepreneur or businessperson who’s kind of going on the web to find this kind of a service. How do they decide which service to go with? Um, and yeah, what, what should they be looking for?

Oleg Kulyk of ScrapingAnt.com: That’s also a really good question. So, uh, one of the competitors I’d really like to highlight, the name of them is, uh, Scraping Bee.

And we are joking about this, that, uh, we have a scraping zoo here, and I also tried to check availability of the main scraping zoo, and it was already, someone already took this domain. So there are like many different websites. Started from

2021 with the name started off, uh, scraping and then the name of some [00:31:00] animal. And yeah, it’s pretty funny, but I’d really like to mention scraping beat those guys from friends. They’re really good. And we recently had a really nice and friendly conversation for the first time. I think that we shouldn’t be a competitor at all.

So sometimes I really understand that use case for some particular business could be better solved by them. And I’m free to suggest them to check out their services as well. Still, the difference that we are provide. Is this one new feature for sure that allows to extract data automatically and pretty easily.

And, uh, also the technology itself allows us to extract data that are unextractable. By other websites. So the way we are extracting this, the way we are [00:32:00] providing and, uh, getting, preparing our browsers, it’s a, in this unique way as well as we are using a huge, I’d say really huge pool of IP addresses close across the globe allows us to.

Extract and get information without getting blocked. And also the thing that could be really useful for businesses that in case of a failure, and in case you’re still getting good blocked for some particular request, you won’t be billed for this one. So we only bill for successful requests, successful data extraction.

Otherwise you don’t need to pay and you’re not. need to spend any API credits, even if you’re at free plan. So we are trying to bring value and it it’s our goal to provide the best service and we are commit doing this by not [00:33:00] billing for our failures. So we are billing for only success.

Sushant Misra of TrepTalks: Is there any, um, what are the, I’m assuming that on the web, if, if, if a data is public on website, it’s public data, anybody can go and get it, but are there any, um, considerations?

I don’t know if it would be considered legal or not, but, um, for a scraping site to go to sites like Amazon, you know, which I’m assuming that probably puts a lot of barriers to prevent scraping on their sites. Is it still considered, you know, anything that’s public is public data. And if I go on your site and figure out a way to, you know, capture that data, it’s, it’s okay.

Like, are there any legal or compliance considerations for, for not just your company, but let’s say your client, you know, who’s, who’s using your site to, [00:34:00] uh, you know, scrape data. Can anybody sue them? Hey, you’re, you know, can Amazon sue someone and say, you know, why are you Coming on my site every day, scraping thousands of pages and collecting data.

Is there any issues there? Or is it that if your data is public, it’s public. You know, I can come and scrape it if I, you know, if I, if I can figure out a way to do that. So.

Oleg Kulyk of ScrapingAnt.com: We are trying to spread the culture of white hat, uh, scraping. So we are trying to provide and suggest the best practices for data extraction that doesn’t.

Really hurt for a website owner and that would not shut like down and like that would not do any malicious actions. [00:35:00] It, uh, we are doing this by monitoring activities. In automatic way that would not allow doing some really something bad, some bad actions using our tool because our main goal is to provide tool for other people to use.

Still, some of the websites could say their privacy policy that it’s not acceptable to use automated access to this particular website. I’m not sure that this is something that. Really makes sense because if you’re providing, uh, your website, uh, if you have an open access to your website, anybody can, uh, uh, go to it and get data out of it.

Uh, the, from another site, it’s possible that some data could [00:36:00] be hidden by a login. So you need to authenticate at some particular website. If you’re using automated. Access to such, uh, parts of the website, it could be not legal. So, we’re trying to explain to our customers that something is right and something is really not.

And they could consider, um, what consequences could be out of some actions. Still, uh, there is a really good… Case, a recent case about Google. So recently they updated their privacy policy for their search services. That if you’re allowing Google to index your website to index your public web pages, you’re also allowing Google to use your data [00:37:00] to train Google’s AI model.

So it leads that if you want to use Google and be present at the Google search, you’re exchanging it to your data and, uh, your data for their model to learn. So it’s one of the recent cases. And the most mentioned case is a real court case between, uh, LinkedIn and some of the web scraping companies. Data, I’d say data extraction companies because they are, they were the initial, they were the initial actors of the extraction process.

So, and LinkedIn tried to sue them, but they failed because all the data they extracted from LinkedIn was publicly available. So, and that’s the case everyone could appeal to. So, if your data is publicly available. [00:38:00] It could be scrapped. Otherwise, you need to protect your data with password or some other protection.

So nobody could access it as well as Google could not index it and scrape it. Is

Sushant Misra of TrepTalks: your tool capable of, uh, um, of, uh, scraping LinkedIn? Can, can somebody just scrape and, and, uh, get. people’s

Oleg Kulyk of ScrapingAnt.com: information? Uh, only the public one. So technically it should be possible with some of the tricks to even log in inside LinkedIn and act on behalf of the profile you log in with.

But yeah, we are providing a fully featured browser capabilities, so yeah, we could get data from LinkedIn and yeah, sometimes even not that [00:39:00] allowed one part. Oh,

Sushant Misra of TrepTalks: that’s awesome. I mean, you know, as you’re talking, I’m thinking of all these different ways that I can use. It’s uh, I mean, data is It’s definitely very useful.

Um, how are you getting new clients? Uh, how are you marketing? So initially you said, you know, you had published your API on a marketplace, but now you’re kind of doing your own things because you can control your marketing and you can, uh, con, you know, have direct connection with your customers, um, what kind of marketing you are doing right now to get new customers and, um, and do you do anything to kind of retain customers or, you know.

Uh, make sure your customers are coming back again.

Oleg Kulyk of ScrapingAnt.com: I think that the best marketing tool is the word of mouse. So, uh, one of our like core values in this company is to provide some value for [00:40:00] others first. So when we are providing some value. To others, we could get, or could not, but still, we are trying to do it this way.

But you can get something back when you are first sharing something. So we are trying to communicate with businesses and help them to innovate their processes by using data. We are trying to help them, and they are spreading this information with them. friendly with their friendly businesses, with their friends, with, uh, like other day sharing feedback at Twitter, how good their processes now.

And the same is applicable for all the content we are trying to share. So, uh, currently I think that Google. It’s for example, Google is capable to understand when some piece of content is just [00:41:00] a CEO based article and when it’s not and when it’s full of some valuable content and helpful content and as we are a technical company, and it could be at Thank It’d be tricky for some companies to integrate with us.

We are trying to share all the tips and tricks and best practices and different ways to how, how to do the data extractions extraction at our website. As we are already. Doing this for like four years and almost a decade. So, and so the marketing, I think, yeah, the mark, the marketing concept of our business is to bring the value using different channels and then get back feedback along with [00:42:00] probably some new customers.

Sushant Misra of TrepTalks: I think, uh. In the AI world, when, you know, writing, creating new content pieces is kind of democratized, right? Anybody can go on chat, GPT, you know, give some prompts and create like a, at least a piece of content, modify it and publish it. I think number one, I think, uh, search engines probably kind of recognize which piece of content is, you know, what percent of its written by human versus AI.

I think over time, maybe more and more value will be placed to original content that’s providing original value rather than just, you know, AI created content. That’s, you know, taking information from the web and it’s really just recreating something because in the AI world, it almost. Disincentivizes people to write original content.

So I think, uh, [00:43:00] anybody who’s creating new ideas, new thoughts, new content, it’s, it becomes even more valuable in the AI world because not everybody has the incentive to do that. Um, what are your thoughts on AI? What are your thoughts on the whole chat GPT thing? Are you, um, do you, do you see it as a positive thing or a negative thing?

And how are you using it right now? So,

Oleg Kulyk of ScrapingAnt.com: like, in short, I think I’m really excited about it. So I was the early adopt adopter of this technology, and this technology allows me to explore this whole world of AI in different angle and. Access the open source, open AI like models. So there are some resources that allows you to have your own model based on [00:44:00] your own needs.

And as you, as we are processing lots of data, we can use this data to teach and to prepare to train our model. So, uh, and we’ve been trying this for the last, I think, two years. I’ve been, so I was really, uh, so when you want to create, to extract some data from the internet, you need to write a parser, the right HTML parser.

And it takes time, and it makes you feel anxious when you need to do it often and for, for different websites. And it’s the same process. You need to investigate the CSS selectors. You need to investigate the best way to. Uh, extract this data, how to process it. And I’ve been looking for the way to automate this instruction for in extraction for every website for [00:45:00] the last two years, but I haven’t seen.

Any technology that would allow me to do it with that is that the modern AI model models do. So I’m really excited. And currently I’m exploring much more information in different niches. I’ve never been aware of using open AI models as it’s like the. Google insight at your fingertips that explains and like, it’s really a good and are curated data at your fingertips in a communicative manner.

So I’m really excited about it and it’s hard to predict how it would evolve, but I think it just a start for this all-new AI world that we would live in, in the next several years.

Sushant Misra of TrepTalks: Um, what does your team look like right now? Are you still only two [00:46:00] people or do you have more people join your team? So we

Oleg Kulyk of ScrapingAnt.com: have two full time founders, me and Andrew, and, uh, as you want to focus on what we want to do, we are outsourcing some of their Uh, business parts to other companies.

So we are sourcing, uh, support and we are outsourcing, uh, accounting, uh, to other people, but. All the technical stuff and sales are done by us.

Sushant Misra of TrepTalks: What is the future vision for your business? What do you see your business? Um, maybe let’s say five years down the road.

Oleg Kulyk of ScrapingAnt.com: So as we started talking about AI, I think that we are going to move in this area fully.

So we are going to provide tooling for companies that are aiming to. [00:47:00] Train their own models or provide our expertise in training models, or even provide some, um, models and something like chat GPT, but, um, that’s related to data extraction, um, niche itself. So, um, I think. From the technical perspective, we would grow into this area as yeah, all the ways of data extraction and working with data won’t work and they are going to pass.

So we need to maintain the pace of this modern world and adopt those technologies. And from the perspective of the business growth, I think the best way to keep our values. As they are now trying to be as small as it possible and to grow [00:48:00] only in the when it’s really needed, but not by just that, okay, we want to hire more.

guys and go out of operations. So we like to do what we are doing and when we are staying small, we can do it. In our own way, and like, yeah, doing it our own pace.

Sushant Misra of TrepTalks: No, that makes a lot of sense. Um. Any, any mistakes or lessons learned since you have, uh, created your business? Um, I know COVID and of course, you know, the war has kind of put a lot of challenges your way, but, you know, just in the sense of business, have you made any mistakes that you kind of regret, or, you know, that has kind of taught you a lesson that you can share with other entrepreneurs? [00:49:00]

Oleg Kulyk of ScrapingAnt.com: So definitely, I like doing mistakes. I think doing mistakes is the best way to learn something and to, to, it’s, it’s the way to success. So, uh, I think the main rule in scope of mistakes is to fail fast, try fast and fail fast because scraping ant is something that I’m, I’m really excited of and I like doing this.

But I’m not sure that is my, that scraping on is my last venture. So doing this business allows me to do as much as possible mistakes as I can do. And I want to do this mistake. I want to try a different approach and understand why they don’t work. So, this way I would be able to be more successful and prevent those mistakes in the future.

So [00:50:00] I think be not that successful, but having a lot of lessons is better than have a blind success.

Sushant Misra of TrepTalks: For sure. Now, I’m going to move on to our rapid-fire segment in this segment. I’m going to ask you a few quick questions and you have to answer them maybe in a one or two words or a sentence or so. So the first one is, uh, a book recommendation for entrepreneurs and why?

Oleg Kulyk of ScrapingAnt.com: Oh, that’s the good question. So best stuff I’ve recently read is the chat GPT output for my prompts, but I’m really excited about podcasts that, um, was. Transistor FM. Those guys are really cool, and their podcast calls build your SAS. So I started listening to their podcast in airport when I was traveling to [00:51:00] our headquarters at Poland and yeah, they are saying.

Something that is outside of the modern common suggestions for, um, I’ve heard from other people. So yeah, I’d suggest listening, they’re podcasted there and they’re cool guys. So I’d suggest this one,

Sushant Misra of TrepTalks: um, an innovative product or idea that you are currently, that you feel excited about, you know, I know you said charge dbt, but any, do you have any other product or idea that you are excited about right now?

Oleg Kulyk of ScrapingAnt.com: All the modern generative AI models are really insanely cool. So meet journey, open AI image generation. They collected all the modern mathematical and technical, uh, innovatives [00:52:00] from the world and. They are publicly available for like for us, for normal consumers. So yeah, it’s, it’s really insane technology.

And I suggest started looking into this and try those technologies. As as fast as you can because being early adopters would be the huge benefit for you for

Sushant Misra of TrepTalks: sure, Uh a business or productivity tool or software that you would recommend or a productivity tip.

Oleg Kulyk of ScrapingAnt.com: I think the best productivity tip is to learn How to rest properly you need to learn how to rest your body and your mind because If you want to be successful, you need to have a good health and good ideas, new good ideas.

And you won’t be able to have a good idea when you are constantly [00:53:00] grinding something. It won’t be good for your health and for your mind. So You have to, despite you want to, that you want to success to come faster, it won’t be that way. So you need to learn how to properly rest and to how to have the best time and not just grinding and be always start.

Sushant Misra of TrepTalks: Yeah, I think, I think that that’s a very, very important for sure. Um, another startup or business that you think is currently doing great things. Oh,

Oleg Kulyk of ScrapingAnt.com: that’s, I think that military tech startups would be a new boom in our world. It’s really sad to say, but that’s true. So there are insane technologies that allows people to save their lives and not be at the front life and to be [00:54:00] replaced by.

robots by drones and I see how this niche is growing in the world now. It’s the basic concept of our future safety, but still, it’s a bit scary that we need to work on this one instead of. But yeah, the military tech is doing great things, a scary one once, but great

Sushant Misra of TrepTalks: things. I mean, that’s so interesting. As you said, it’s, it’s sad that we have to work on these things.

I think. Given that everything that we know now, um, I think the all the nations of the globe should be collectively thinking about peace. I think we should be thinking about military tech to to save ourselves from alien invasion. And, and maybe also like from catastrophe, like, you know, uh, [00:55:00] asteroid or meteor, uh, uh, colliding with earth or something like how to divert that.

So I think creating technologies for those things, rather than fighting with ourselves, I think that’s probably a better thing. Um, A peer entrepreneur or business person whom you look up to or someone who inspires you? So

Oleg Kulyk of ScrapingAnt.com: yeah, there are some really well-known entrepreneurs and owners of the media platforms.

They are not great people, so I won’t, I don’t want to mention them, but I really want to mention my parents that I’m really, I’m really proud of the work they did in the past and how they transformed their lives. And yeah, I, they are like, I’m trying to be as that good as they been like back in days. [00:56:00]

Sushant Misra of TrepTalks: Awesome. Final question. Best business advice that you have ever received, or you would give to other entrepreneurs?

Oleg Kulyk of ScrapingAnt.com: So, yeah, I think the best advice that it would take. Way like it would take much longer than you expect. So you have to choose a right thing to do. You need to do something that you really like because doing something that you don’t like for a long time, one brings you success, despite some niche or despite some trends.

So just find something that you really like to do and You won’t even like, you won’t even be, you won’t even like think bad things or you won’t have, have low morale regarding something if you really like to do something and you would get, you will have the success if you do [00:57:00] the thing that you really like.

Sushant Misra of TrepTalks: I think, I think that’s a great advice. I think if you’re Generally passionate about something you’ll, you know, you’re doing it in a way because you like to do it, you know, besides, despite, you know, setbacks, failures, challenges, uh, you’ll continue moving forward. And I think that’s very important. Uh, and I think the other thing that you said, you know, it’s going to take more time than you’ve anticipated.

I think that’s, that’s very, very true. Well, Oleg, those were all the questions that I had. Thank you so much again for joining me today. Um, if anybody wants to check out your product or getting. Get in touch with you, what is the best way to do that? So you can

Oleg Kulyk of ScrapingAnt.com: visit our website, you can just, uh, check Google and uh, search for scraping or go to scraping and.com.

And yeah, you can reach me at Twitter. We are our official Twitter profile, scraping and, and using our support mail. So yeah. [00:58:00] Just write something like I want to contact Oleg and your letter would come to me.

Sushant Misra of TrepTalks: Awesome. Well, Oleg, thank you so much again for sharing your story, for sharing, uh, you know, your lessons learned entrepreneurial and also personal.

And, uh, yeah, thank you so much again for joining me today at TrepTalks. Wish you all the

Oleg Kulyk of ScrapingAnt.com: very best. Thank you, Sushant. It was really, really great to talk with you. Thank you.